Oracle and OpenAI Skipped the Financing Question Investors Keep Asking

Oracle and OpenAI pitched ChatGPT and Codex inside Oracle’s software at Oracle AI World. Neither CEO mentioned how Oracle funds its $664 billion contract backlog.

Oracle co-CEO Clay Magouyrk and OpenAI CEO Sam Altman kept the focus on adoption. Shareholders, however, are weighing a backlog about half tied to OpenAI against negative free cash flow.

Is Oracle Selling AI Software or Renting AI Capacity?

At the keynote, Altman said GPT-6 drew OpenAI’s strongest adoption of any new model. He cited coding, enterprise software, security, and science.

Magouyrk said AI now writes more than 70% of Oracle’s own software.

Meanwhile, OpenAI’s case study counts 130,000 active ChatGPT users inside Oracle. It also counts more than 95,000 users of Codex, OpenAI’s coding agent.

The same study says Oracle’s talent acquisition team cut research time by 98%.

Oracle said Tuesday the additions are meant to make databases more actionable, CNBC reported. That goal could shift revenue toward software and away from rented capacity.

Why Did the Oracle and OpenAI Pitch Skip Oracle’s Financing?

The on-stage conversation included no figures on capital spending, backlog, or financing.

Oracle reported negative free cash flow of $5 billion for its fiscal first quarter. CNBC reported capital spending hit $28.5 billion in that period, up from $8.5 billion a year earlier. The company also completed a $20 billion common stock sale during the quarter.

In contrast, Magouyrk told analysts Oracle closed more than $30 billion of new AI contracts without needing additional capital.

Backlog, meaning contracted revenue not yet recognized, reached $664 billion. MarketWise puts OpenAI at about half of that.

Oracle will also offer Jalapeño, an OpenAI chip built to run trained AI models, on Oracle Cloud Infrastructure (OCI).

If OpenAI’s demand holds, the backlog converts to revenue and the silence matters less. If demand slows, the concentration becomes the story.

The gap reflects a split across the AI trade. Model makers sell adoption, while infrastructure providers answer to lenders. Oracle straddles both, so software optimism holds only if OpenAI keeps paying for the capacity.